Although the link between maternal and child depression is well established, little is known about the mechanisms by which this risk is conferred. The proposed study is designed to address this gap. In so doing, we seek to integrate and extend findings from two separate lines of research: cognitive-interpersonal models of depression and psychiatric genetics. The proposed study involves a longitudinal investigation of mothers and their children (ages 8-14 years) drawn from the community and assessed every six months for two years. Two groups of mothers will be recruited for the study: (a) those with a history of major depressive disorder (MDD) during their child's life and (b) those with no lifetime history of any depressive disorder and no current Axis I disorder. Our primary goal is to examine the development of children's cognitive vulnerability to depression (negative attributional style and attention and interpretation biases for facial displays of emotion). We hypothesize that maternal history of MDD will be associated with the presence of more negative cognitive styles in their children and that these cognitive styles will become more negative over the course of the follow-up as a function of mother's ongoing depressive symptom levels. We also hypothesize that negative events in the children's lives (maternal criticism as well as general negative events) will predict negative changes in children's cognitive styles. Integrating recent findings from psychiatric genetics, we predict that these effects on children's cognitive styles will be moderated by two specific genetic risk factors - a functional polymorphism in the serotonin transporter gene (5-HTTLPR) and a candidate polymorphism in the brain-derived neurotrophic factor (BDNF) gene. Our secondary goal is to test a transactional cognitive-interpersonal mediation model linking maternal and child depression. In testing this model, we hypothesize that mothers with a history of MDD, compared to control mothers, will exhibit elevated levels of depressive symptoms across the follow-up. These depressive symptom elevations are hypothesized to contribute to the occurrence of negative events in children's lives, which will then contribute to the development of children's negative cognitive styles. These negative cognitive styles, then, are hypothesized to contribute to children's symptoms and diagnoses of depression. Building from the stress-generation literature, we will also test the hypothesis that children's levels of depression will reciprocally predict prospective changes in negative events and children's cognitive styles. We also predict that children's 5-HTTLPR and BDNF genotypes will moderate the impact of mother depression, negative events, and children's cognitions upon child depression, allowing increased precision in predicting which children are at greatest risk for the intergenerational transmission of depression. A tertiary goal is to examine the specificity of the links in the mediational model to children's depression versus other disorders (e.g., anxiety and disruptive behavior disorders).